Ilia Kulikov

LLM Pretraining with Continuous Concepts

J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learning

Hybrid Reinforcement: when reward is sparse, better to be dense

OptimalThinkingBench: Evaluating Over and Underthinking in LLMs

The Majority is not always right: RL training for solution aggregation

Learning to Reason for Factuality

Adaptive Decoding via Latent Preference Optimization

Bridging Offline and Online Reinforcement Learning for LLMs

Diverse Preference Optimization

CoT-Self-Instruct: Building high-quality synthetic prompts data for reasoning and non-reasoning tasks

Self-Taught Evaluators